Internet Censors Must Be Accountable For The Things They Break

Yesterday's scandal over the UK Internet Watch Foundation's attempt to censor a purportedly pedophiliac Wikipedia entry raises some important questions about unintended technical consequences of Internet censorship systems. The Wikipedia article was about a 1970s hard rock album called Virgin Killer [NSFW] by the German band Scorpions.

Censorship technologies are purveyed as a way to protect us from the evils of child abuse. But they're costly systems that are unlikely to actually protect anyone or prevent any child abuse — they're more likely to interfere with the way the Internet works and hamper innovation by online communities.

The Wikipedia entry for Virgin Killer contains an image of the album's cover, which depicts a nude teenage girl, with only minimal obscurement by a "shattered glass" effect. The article contains an encyclopedic discussion of how executives at RCA Records had decided on the cover concept, the controversy it had caused at the time, and how the woman depicted had no objections to the image's use, even in retrospect. Wikipedia had come under pressure from conservative Christian groups in the United States to remove the image, but through extensive debate Wikipedia editors agreed to keep the image with the article.

Last weekend, the Internet Watch Foundation (IWF), a coalition of UK ISPs, Internet companies and censorware vendors, listed the image as a "potentially illegal indecent image of a child hosted outside the UK". This caused a number of UK ISPs to begin routing all Wikipedia traffic through filtering proxy servers, which disallowed access not only to the cover image but the entire Virgin Killer article.

This is where the unintended consequences began. Some of the ISPs' proxy servers were unable to handle the volume of traffic Wikipedia attracts, making Wikipedia unavailable entirely. But there were also stranger side effects.

Wikipedia is an encyclopedia which allows anyone to edit an article by clicking an "edit" button, changing the text, and clicking "save". Because of the site's enormous visibility, there are thousands of people around the planet who are continuously trying to deface pages by editing them. Wikipedia responds to the most persistent attempts at vandalism by blacklisting IP addresses that are repeatedly the source of vandalistic edits. In some cases, users can get around the blocks if they have accounts at Wikipedia, but in severe cases, users must petition to have their account allowed through the block. Without these blocks, the encyclopedia would have long ago been reduced to gibberish.

What does this have to do with censorship by the IWF? It turns out that IWF had decided upon a method for censorship which was incompatible with Wikipedia's anti-vandalism architecture. The proxy servers that IWF member ISPs used to block the Virgin Killer article meant that Internet users at Britain's largest ISPs were suddenly sharing a handful of proxy servers. Many of those proxy servers were configured to make the users affected by them indistinguishable to Wikipedia. Enough of those indistinguishable users were vandalizing Wikipedia that site admins were forced to block all of them from editing.

That meant that huge numbers of ordinary British Internet users could no longer edit Wikipedia, because technical decisions by Internet censors suddenly caused them to be sharing IP addresses with a horde of vandals.

Today, the IWF reversed its previous ban on the Virgin Killer article, concluding that the intervention had been counterproductive because it was causing more people to see the allegedly pornographic image. We agree with their decision, but they have the wrong reasoning: they had no business censoring that article in the first place — the community of Wikipedia editors is if anything the more legitimate, reliable and grown-up adjudicator of which images are appropriate subject matter for an encyclopedia. And the block on the editing of Wikipedia by a large portion of the UK population just highlights the dangers of deleterious unintended consequences once we travel down the low road of Internet censorship.

The IWF's censorship failed doubly to be transparent. It is not transparent technologically, which leads inevitably to a conflict the end-to-end expectations of the rest of the Net. And its process of attempting to block and filter is far from transparent to those who are caught up in it: from puzzled and frustrated Wikipedia users to the millions of Britons who never realized they were paying their ISPs for a compulsorily and arbitrarily sanitized Internet.

Update, 10th of December: Mark Pellegrini, Wikimedia Foundation Communications Committee member, writes to point out that the IWF blacklist only blocked the Virgin Killer article and the image's description page; they did not block the actual URL of the controversial image itself.

[Permalink]

Change.gov Content Now Under Creative Commons License

In the last few days, President-elect Obama's transition team took a significant stride towards a more open government by licensing the content of Change.gov under a Creative Commons Attribution license. Using that license essentially means that the transition team is allowing others to freely share and remix what's posted there, provided that reposts are attributed to Change.gov. The move is a victory for the public and the many advocates for a more wired, participatory democracy.

It's also another reminder of the importance of Creative Commons, which affords creators an opportunity to opt for something less than Disney-style copyright restrictions. By embracing a CC license, the Obama team sets a valuable example for others in government, many of whom may have defaulted to "all rights reserved" without considering other options.

While Change.gov has experienced some growing pains, the transition team appears to be making a real effort to use the website as a legitimate location for its conversation with the American public. The preview post of the President-elect's planned weekly address (posted on Thanksgiving Day) includes links to multiple sources — an embedded YouTube video, a link to the same video posted to Yahoo! Video, and a high-resolution .mov file — with the Creative Commons license guaranteeing that the public can freely share, remix, comment, and report on the President-elect's statement.

The switch to Creative Commons licensing is encouraging and we hope that it is a herald of more pro-open government changes to come.

[Permalink]

Censorship in the 21st Century: Targeting Intermediaries

On November 12, 2008, a group of artists and activists unveiled a brilliant spoof of the New York Times, widely distributed to readers in New York and Los Angeles. This "July 4, 2009" version of the Times — which the real New York Times described as a "Grade-A caper" — boldly announced the end of the Iraq War, the nationalization of major oil conglomerates, the elimination of tuition at public universities, and the indictment of soon-to-be-former president Bush on charges of high treason. The poignant send-up, also available in an online version at www.nytimes-se.com, is a perfect example of parody in the 21st century. It certainly got its fair share of attention.

Could the lawyers be far behind? Not surprisingly, the corporate targets of the parody were not pleased. Now, in what is becoming an all-too-familiar trend, one of those corporations has attempted to shut down the site by putting pressure on what is often the weakest link in the online speech chain: the domain name registrar. Stymied by the First Amendment and other legal impediments, those who don't appreciate critical commentary and other "objectionable" online content have found intermediaries — providers of indispensable technical services like domain name registration and web hosting — much easier to intimidate.

This time, the complaining (and overreaching) party was the South African diamond conglomerate De Beers, the target of a critical fake ad on the web version of the New York Times spoof announcing that diamond purchases "will enable us to donate a prosthetic for an African whose hand was lost in diamond conflicts." Miffed by the criticism, De Beers responded not by confronting the authors (whose parody is protected by the First Amendment) but instead by threatening their Swiss-based domain name registrar, Joker.com. De Beers has demanded that Joker.com disable the spoof website's domain name or face liability for trademark infringement.

This certainly isn't the first time that a sore target of criticism has threatened an internet intermediary in order to take down content it didn't like. The motivation is simple: all too often, the intermediary will prove to be unwilling to stand up for the rights of its customers. For example:

• In September, 2008, the state of Kentucky initiated an ex parte proceeding aimed at seizing 141 domain names pointing to overseas websites permitting online gambling, arguing that the domain names were illegal "gambling devices" under Kentucky law. Without giving the domain name owners an adequate opportunity to defend themselves, the trial court ordered the registrars with control over the targeted domain names to transfer them to the state, a decision the court affirmed in October. Upon receipt of the court's order, some of the registrars (such as GoDaddy) who were likely outside of the court's jurisdiction nonetheless locked the domain names and purportedly transferred control to the Kentucky court. (EFF filed an amicus brief arguing that the orders were unconstitutional; the matter is pending appeal.)
• In January, 2008, Swiss bank Julius Baer filed suit against (among others) Dynadot, the domain name registrar with which the domain name wikileaks.org was registered. The Wikileaks website to which the domain name pointed — a self-styled "uncensorable Wikipedia for untraceable mass document leaking and analysis" — hosted (and still hosts) documents which anonymous third-party posters argue document financial wrongdoing. Instead of challenging the operators of the Wikileaks site directly, Julius Baer targeted Dynadot which quickly agreed to lock and permanently disable its customer's domain name. (In response to briefs such as the one filed by EFF pointing out the illegality of the court's order approving of the ill-advised agreement, the district court ultimately dissolved the permanent injunction, leading Julius Baer to dismiss its case.)
• In December, 2006, on behalf of its affiliate radio station KSFO-AM, media giant ABC sent a cease and desist letter to 1 & 1 Internet, the then-host of the blog www.spockosbrain.com, a site that criticized what its author argued was offensive and violent rhetoric broadcast by KSFO. Despite the fact that it had no risk of liability, 1 & 1 Internet promptly shut down Spocko's site in response to receiving the threatening letter. (EFF subsequently represented Spocko, successfully moving him to Computer Tyme, a web hosting company that promised to stand up to future threats.)

Intermediaries frequently argue (incorrectly) that they have no choice but to shut down domain names or sites when they receive a legal complaint. But they do have a choice, at least in the United States. As EFF has explained to De Beers, U.S. law provides ample protection for intermediaries, in large part to ensure that online speech and commerce continues to thrive. For example, in 1996, Congress passed Section 230 of the Communications Decency Act which immunizes internet intermediaries from most kinds of liability associated with the content that their customers place on their own sites. Similarly, the Digital Millennium Copyright Act (passed in 1998) provides a safe harbor from copyright infringement liability for intermediaries who follow straightforward procedures in response to valid takedown notices. With these strong legal protections, intermediaries can and should refuse to respond to pressure from companies like De Beers.

Will the poignant work of the New York Times parodists stay up for the world to see? Will other critical online speakers be silenced by improper threats against their virtual soapboxes? The answer may ultimately depend more on the ability of targets to intimidate domain name registrars than on the legality of the underlying speech itself.

[Permalink]

DMCA: Ten Years of Unintended Consequences

Today is the tenth anniversary of the Digital Millennium Copyright Act (DMCA), signed into law by President Bill Clinton on October 28, 1998. EFF is marking the occasion with the release of a 19-page report that focuses on the most notorious part of the law: the ban on "circumventing" digital rights management (DRM) and other "technological protection measures." The report, entitled Unintended Consequences: Ten Years Under the DMCA, collects reported cases where the DMCA was used not against copyright infringers, but instead against consumers, scientists and legitimate competitors.

The collected stories are like a trip down memory lane for those who have followed digital freedom issues over the past decade. Here are a few examples of DMCA abuse in the report that you might remember:

• In 1999, Sony sues Connectix over the Virtual Game Station, which let you play your legit Playstation games on your Macintosh.
• In 2001, the Secure Digital Music Initiative (SDMI) threatens Princeton Professor Ed Felten's research team over disclosure of vulnerabilities in audio watermarking technology.
• In 2001, Russian programmer Dmitry Sklyarov is arrested after speaking at Defcon, accused of building software for his employer, ElcomSoft, that converted Adobe e-books to PDF.
• In 2002, Blizzard sues a group of hobbyist open source developers over bnetd, server software that allows people to play Blizzard games against each other over the Internet.
• In 2003, Lexmark uses the DMCA to block distribution of chips that allow refilling of laser toner cartridges.
• In 2004, Hollywood succeeds in shutting down 321 Studios' DVD X Copy software, which allowed people to make backup copies of their own DVDs.
• In 2006, computer security researchers at Princeton delay disclosure of the Sony-BMG "rootkit" based on fears of DMCA liability.
• In 2008, Hollywood targets Real Networks over RealDVD, software that allows you to copy DVDs to a hard drive for later viewing.

The collection of stories makes vividly clear what EFF has been saying for the past ten years: the DMCA has harmed fair use, free speech, scientific research, and legitimate competition.

That's all the more galling because the law has failed in its stated goal of preventing digital piracy, instead being used to prop up weak DRM schemes whose only purpose is to hinder competition, innovation, and interoperability. That explains why the music industry has largely abandoned DRM, while the Hollywood studios cling to it more fervently than ever.

Not everything in the DMCA is bad. While the anti-circumvention provisions have proven to be a dangerous failure, the so-called "safe harbor" provisions for online service providers have succeeded in creating enough legal certainty to launch companies like Yahoo, Google, eBay, YouTube, and MySpace. Of course, copyright owners have been working hard in cases like Viacom v. YouTube and Io v. Veoh to erode these safe harbors. And, while the safe harbors have protected intermediaries like Google, they have not adequately protected the free speech interests of internet users, as the McCain-Palin campaign recently learned.

There have been recent rumors that the new Congress might reopen the DMCA, creating an opportunity for reform. Unfortunately, that may also create an opportunity for MPAA and RIAA mischief. For now, here's hoping that the DRM continues its slow death and the anti-circumvention provisions become less relevant to real businesses, while the courts continue to interpret the safe harbors to leave a door open to the Internet's disruptive innovators.

P.S. For more perspectives on the DMCA's origins and legacy during this 10 year anniversary week, see Freedom to Tinker and the Public Knowledge blog all this week.

[Permalink]

YouTube Responds to McCain Campaign's Letter

Yesterday, we wrote about the McCain-Palin campaign's letter to YouTube, highlighting how DMCA takedown notices can make online speech disappear from the Internet, even when the claims of infringement plainly lack any merit.

Today, we bring you YouTube's response. YouTube's response points out, much like we did yesterday, that the McCain-Palin campaign's proposed solution (human review of DMCA takedown notices targeting videos posted by political candidates and campaigns) favors speech from one particular class of users. YouTube says that it "tri[es] to be careful not to favor one category of content on [its] site over others, and to treat all of [its] users fairly, regardless of whether they are an individual, a large corporation, or a candidate for public office."

At the end of the day, we agree with YouTube that "[t]he real problem here is individuals and entities that abuse the DMCA takedown process." And we commend YouTube for taking action in some cases where it has identified false takedown notices.

Nonetheless, although YouTube may not be the source of the problem, that doesn't mean it can't do more to be part of the solution. YouTube notes that it can't always be certain whether a video qualifies as fair use, and that it can't know whether the poster has a license to the content. That's all true.

But just because YouTube can't always identify sham takedown notices doesn't mean it can't sometimes know the answer. Using a short excerpt from a news broadcast and commenting on it in a political commercial is clearly fair use. And there are many other examples of clear fair uses, as well.

We'd love to see YouTube take further action, so that takedown notices directed at clearly non-infringing videos can't be used to silence speech. As we said yesterday, stay tuned for more on this topic from us soon.

[Permalink]

McCain Campaign Feels DMCA Sting

Yesterday, the McCain-Palin campaign sent a letter to YouTube describing the troubles it has been having with bogus DMCA takedowns targeting its videos:

[O]verreaching copyright claims have resulted in the removal of non-infringing campaign videos from YouTube, thus silencing political speech. Numerous times during the course of the campaign, our advertisements or web videos have been the subject of DMCA takedown notices regarding uses that are clearly privileged under the fair use doctrine. The uses at issue have been the inclusion of fewer than ten seconds of footage from news broadcasts in campaign ads or videos, as a basis for commentary on the issues presented in the news reports, or on the reports themselves. These are paradigmatic examples of fair use...

It's heartening to see a presidential campaign recognize the importance of fair use and "remix culture" (the Obama-Biden campaign has also been the victim of frivolous takedowns from big media companies, so this is a bipartisan problem). EFF, the ACLU, Harvard's Citizen's Media Law Project, and Stanford's Fair Use Project have been making the same point for several years now. EFF has also been providing direct legal assistance to victims of DMCA abuse.

Unfortunately, the solution proposed by the McCain campaign addresses only the tip of the iceberg:

[W]e believe that it would consume few resources--and provide enormous benefit--for YouTube to commit to a full legal review of all takedown notices on videos posted from accounts controlled by (at least) political candidates and campaigns.

The obvious problem with this solution? It assumes that YouTube should prioritize the campaigns' fair use rights, rather than those of the rest of us. That seems precisely backwards, since the most exciting new possibilities on YouTube are for amateur political expression by the voters themselves. After all, the campaigns have no trouble getting the same ads out on television and radio, options not available to most YouTubers.

Let's start by identifying the real villains here: the major news media outlets. They are the ones censoring these political ads, based on the use of a few seconds of their footage. The networks need to back off and give fair use a wide berth. So let's start by shaming the bad guys here. In addition, lawsuits might help. Under the DMCA, both the campaigns themselves and YouTube have standing to sue those who send clearly bogus takedown notices. (EFF has represented video creators in a number of these cases, including against Viacom.)

There are other possible solutions, as well. Stay tuned for our specific ideas on what YouTube can do to protect fair use while staying within the bounds of its DMCA safe harbor protection (hint: as the McCain-Palin letter points out, you don't need a safe harbor if the video isn't infringing, something that human review by YouTube should be able to determine).

UPDATE: The McCain-Palin campaign has identified the news outlets behind the YouTube removals: CBS, Fox News, and the Christian Broadcasting Network. We noted above that NBC has targeted an Obama-Biden video for removal. That's four news entities that should know better.

[Permalink]

EFF and ACLU of Northern California to ISPs and Content Owners: Do Your Part to Protect Political Speech

Coauthored by ACLU of Northern California Technology and Civil Liberties Policy Director Nicole A. Ozer

On blogs, personal and political websites, and through user generated content sites, ordinary citizens in extraordinary numbers are recreating a public sphere and reinvigorating the democratic debate at the core of our political system. 46% of Americans have already used the Internet in connection with the political campaign- more than during all of 2004.1 User-generated content is playing a particularly integral role, with 35% of Americans watching online videos and 10% using social networking sites to engage in political activity. 2

An overwhelming number of political discussions are taking place in publicly-accessible but privately-owned, online town squares. Which means that this important political speech depends on service providers, users, and content owners all doing their part to safeguard free speech.

Unfortunately, political speech has been threatened repeatedly by claims that controversial material violates a site’s terms of use or infringes copyrights or trademark rights. Here are just a few recent examples:

• The International Olympics Committee demanded that YouTube remove a video of a protest by Students For A Free Tibet, based on a bogus copyright infringement claim. The IOC subsequently withdrew the notice, but the IOC’s demand is a lesson in the dangers of hair-trigger DMCA takedowns by service providers.
• An alleged terms of service violation caused YouTube to take down a slideshow of a military funeral.
• Another alleged terms of use violation caused YouTube to remove a video critical of John McCain, apparently because the video included numerous graphic images of the effects of war. But those images were integral to the commentary: the video focused on McCain’s support for the Iraq war.
• An apparent copyright complaint caused Broadview Networks to shut down a political website parodying Exxon’s environmental policies.
• The Republican National Committee threatened the online vendor CafePress for allowing users to create t-shirts using Republican trademarks, like "Grand Old Party," or the official version of the elephant logo.
• The Chicago Auto Show tried to use allegations of trademark infringement to force the shutdown of a satirical website promoting transportation alternatives.
• The Associated Press tried to use the Digital Millennium Copyright Act (DMCA) to force the takedown of blog entries that reproduced excerpts of AP news stories—some of them just a few words long.

ACLU of Northern California and EFF urge service providers to take extra precautions before pulling the plug on political speech. Remember that you're facilitating a new era of reason and debate, and that there are laws that protect you as a facilitator. By taking that responsibility seriously, you’ll do right by your users, content owners and the political process.

We urge content owners to count to ten and look at the Fair Use Frequently Asked Questions and Fair Use Principles for User-Generated Video Content for some guidance before firing off a complaint. Remember that you are legally obligated to consider whether the use of your material is a fair use. Consider carefully whether actions may result in the loss of free speech, and remember: the antidote to free speech that you don't like is MORE free speech. Make your voice heard with a written blog post, a video blog post, or a message in the comment thread. We also urge users to contact us if they feel that their political speech has been improperly censored.

As we move forward into the fall election season, the Internet can continue to revitalize our political lives in exciting and unforeseen ways—but only if service providers, users and content owners all do their parts. No matter where you stand on the candidates or the issues, we should all agree on one principle: No downtime for online free speech!

1. 1. https://www.pewinternet.org/pdfs/PIP_2008_election.pdf
2. 2. Id. As many as 65-milliion Americans watched the virally distributed, political web video, “This Land,” in 2004. Under the Radar and Over the Top: Online Political Videos in the 2004 Election. https://www.ipdi.org/UploadedFiles/web_videos.pdf

[Permalink]

Wikipedia Wins Dismissal of Baseless Defamation Claims

In a victory for free speech and user-generated content, a New Jersey judge has dismissed baseless defamation claims against the operator of Wikipedia. In a recent ruling, the judge correctly found that federal law immunizes the Wikimedia Foundation from liability for statements made by its users.

This case began when literary agent Barbara Bauer sued Wikimedia, claiming the organization was liable for statements identifying her as one of the "dumbest of the twenty worst" agents and that she had "no documented sales at all." EFF and the law firm of Sheppard Mullin Richter & Hampton represented Wikimedia, and moved to dismiss the case in May, arguing that under Section 230 of the Communications Decency Act, operators of "interactive computer services" such as Wikipedia cannot be held liable for users' comments.

Section 230's blanket protection of sites like Wikipedia does not mean that alleged defamation on the Internet cannot be challenged in court. Instead, the law requires that litigants direct their efforts at the speakers themselves and not the forums where statements were made. This means that sites like Wikipedia, Craigslist, and other online communities can continue to include user-generated content without living in fear of costly lawsuits.

Since it was signed into law over a decade ago, courts across the country have consistently applied the protections of Section 230 broadly, fulfilling Congress' intent to foster online interactive communications. We're pleased to add this case to the list.

[Permalink]

In Memoriam: Ed Foster, 1949-2008

Ed Foster, the journalist and consumer advocate behind InfoWorld's GripeLine column and GripeLog blog, died of a heart attack this weekend. He was 59.

It's no exaggeration to say that Ed was one of the preeminent consumer rights activists of the digital age. During his more than 20 years as a "reader advocate" at InfoWorld, he was far ahead of his time, recognizing that in a world increasingly dominated by software and online services, the digital consumer needed a champion when squaring off against the likes of Microsoft, Adobe or AutoDesk. Following in the traditions of the best consumer reporters before him, Ed exposed software vendors and online service providers that treated their customers shabbily.

But it was in his tireless work against "sneakwraps" -- those "end user license agreements" (EULAs) and "terms of service" (TOS) that require our "agreement" -- that Ed was without peer. You may not be reading all those "agreements" before you click thru, but Ed was. He recognized earlier than most that sneakwraps were going to be the digital consumer's worst nemesis, the mechanism that stripped consumers of the legal protections they enjoy when buying a book, a chair, or an automobile. Long before most consumer groups were thinking about sneakwraps, Ed was covering and participating in efforts to block UCITA, a package of state laws pushed by large software vendors that would have stripped consumers of valuable protections under contract law (UCITA was ultimately adopted by only two states, VA and MD, and has since been abandoned). Ed also contributed his insights on DRM, product activation, and reverse engineering to groups like AFFECT (Americans For Fair Electronic Tranactions) and EFF, making sure we knew what consumers were dealing with in the trenches.

Ed will be sorely missed, both professionally and personally, by all who benefited from his wisdom. Here are a few of my personal favorites from among his remarkable output of columns and posts:

Embroidering on a Copyright Shakedown Theme -- casting the spotlight on the "Embroidery Software Protection Coalition" (ESPC) after it sent settlement demand letters to grandmothers who bought embroidery software on eBay. Based on Ed's tip, EFF stepped in to protect the interests of innocent purchasers.

Sneakwrap Files: McAfee Automatic Renewals -- a consumer advocate's classic, wherein Ed confronts McAfee over the "automatic renewal" provision buried in the fine print of their EULA. McAfee backs down and coughs up a refund.

The Lexmark Car -- an April Fool's post explaining what the world would be like if a car manufacturer tried to get away with the kinds of shenanigans practiced by Lexmark in connection with their laser toner cartridges.

Into the DMCA Groove -- in 2003, an eBay seller gets into hot water after trying to auction a promo CD given away at The Gap. Ed cries foul, predicting 5 years in advance the exact outcome of EFF's UMG v. Augusto case, where a court found that "promo use only" labels can't trump the first sale doctrine.

[Permalink]

More ISPs Decide to Filter Usenet Newsgroups

New York Attorney General Andrew Cuomo recently succeeding in pressuring AOL and AT&T to join the ranks of Verizon, Sprint, and Time Warner Cable in limiting access to many or all of the Usenet newsgroups hosted on their servers. This tactic will hinder free speech and the access to information in Usenet communities, without deterring the child pornographer. But since the ISPs are “voluntarily” bowing to political pressure, rather than obeying a statutory edict, traditional First Amendment court challenges are unlikely to protect these online communities.

Attorney General Cuomo has pressured these companies into censoring enormous amounts of First Amendment-protected material after an investigation found 88 groups containing child pornography, or 0.5% of the active discussion groups in the alt.* hierarchy. Verizon and Sprint are taking down one gigantic subset of groups, the very popular alt.* hierarchy, AT&T will block all alt.binaries.* groups, while Time Warner Cable and AOL are shutting down their Usenet service entirely. (link) California's Governor and Attorney General have jointly called on California's service providers to follow New York's initiative in "removing child pornography from existing servers and blocking channels that disseminate the illegal material."

Usenet is a technology that predates the birth of the World Wide Web. Similar to bulletin boards, it allows for conversation threads on a wide variety of topics. The alt.* hierarchy alone contains nearly 19,000 different groups. (link) Blocking the alt.* hierarchy, which is the largest and most active, predominantly censors innocent discussions in order to stop illicit activity in a handful of them. Examples of some of the groups that will no longer be available are alt.adoption, alt.culture, and even alt.horology (discussing the science of timekeeping). Blocking all of Usenet throws out even more legitimate and useful expressive speech.

Congress and the courts have struggled for more than 10 years to address the issue of "objectionable" Internet content, without a constitutionally permissible result. Two well-known attempts by Congress were the passage of the 1996 Communications Decency Act and the 1998 Child Online Protection Act. Both attempted to punish individuals who transmitted indecent material that was harmful to minors. The U.S. Supreme Court has ruled these provisions unconstitutional because they block speech that would be protected by the First Amendment in contexts outside of the Internet. States have also attempted to preemptively censor material. In 2002, the Pennsylvania legislature attempted to hold ISPs criminally liable for child pornography available on the Internet. The Pennsylvania Attorney General was able to unilaterally and secretly order ISPs to block access to IP addresses suspected of containing child pornography, resulting in the blocking of many innocent sites (particularly when the same IP address was used to host a variety of websites). A state court soon struck down that law as unconstitutional for violating both due process and prior restraint of speech.

The censorship demanded by NY Attorney General Cuomo arbitrarily filters an entire electronic neighborhood due to the activities of a few of its residents. Measures already existed to take down objectionable material from Usenet. The kind of responsive enforcement that has been previously used balances free speech and the need to stop the dissemination of child pornography much better than the arbitrary, blanket ban to which the ISPs have agreed. Even so, the tactic will not stop committed child pornographers. ISPs are only blocking their own internal Usenet servers, access to Usenet groups on remote websites remains unblocked. As a result, these actions will do little to stop the threat they are intended to prevent. (link)

Because imposed online censorship inevitably goes to far and has never been upheld in the courts, it is not surprising that Attorney General Cuomo chose to apply political rather than legal pressure to the ISPs. While Internet users have First Amendment rights to speech and expression online, ISPs also have a First Amendment right not to carry content that they do not want to promote. To the extent the filtering program is “voluntary”, it is not the government censoring Usenet or the ISPs, but the ISPs choosing what speech they want to carry. Still, ISPs offer the critical public service of providing people with access to the prosperity of information available online. They have an obligation to consider the effects of their censorship on innocent material, and not just the political expediency of accommodating the Attorney General’s request. The heavy-handed approach of shutting down an entire online service is the least efficient and effective way to accomplish their legitimate goal. Usenet is a communication protocol, no different from Hypertext Transfer Protocol (HTTP), which makes the Web possible. Although the Web is also plagued by some amount of objectionable material, no one would remotely consider blocking access to the entire internet an appropriate solution. ISPs should not take advantage of the relatively small size of the Usenet community to deny their customers access to the diverse forms that content can take on digital networks.

*Thanks to Nick Jackson, summer intern, for his work on this Deeplink

[Permalink]